Modelagem e previsão de volatilidade determinística e estocástica para a série do Ibovespa

The variance of an asset is the most important information for an investor that deals inthe financial markets. The measurement of that volatility can be done in two differentways. The first one, deterministic case, is done by taking as a starting point theknowledge of conditional variance. In the other approach, called stochastic volatility,one does not know a priori the volatility of the asset. These models are used in manydifferent formulations to explain the specific characteristics observed in the financialtime series, such as volatility clustering, leverage effect and persistence of the volatility.In this paper, the volatility of the Sdo Paulo Stock Exchange Index (Ibovespa) is modelledby the two processes described above. The period analysed goes from july/94 untiloctober/98, include three critical periods of the world financial markets: the Mexicancrisis, the Asian crisis and finally the Russian debacle. The main conclusion is that both process perform quite well.